[ThePath]

EDIT: replaced 1024x760 with 1080p benchmarks



Source

So, what people think about this ?

Test is good for AMD -> Test is valid ?
Test is good for Intel -> Test is wrong ?

 

[Cloudfire777]

1024x768?

 

[Kasp1js]

Quote:

Originally Posted by Cloudfire777

1024x768?

True dat, Intel almost always wins at synthetic benches.

 

[jprovido]

Quote:

Originally Posted by Kasp1js

True dat, Intel almost always wins at synthetic benches.

interestingly enough amd handily beats intel cpu's on the "welcome to the jungle" benchmark.

 

[White Fire]

This is virtually useless info.

Nobody judges the game based on it performance at 1024x768

 

[ThePath]

Quote:

Originally Posted by Cloudfire777

1024x768?

In the other thread which AMD was winning, it was tested @ 1280x720, and yet AMD fanboys did not say that it was useless test for some reason.
http://www.overclock.net/t/1363058/pc-games-hardware-crysis-3-in-the-cpu-test-amd-fx-processors-dominate-our-benchmarks

So, 1024x768 is low resolution but 1280x720 is high ?

It also worth to mention that in other benchmark they used "welcome to the jungle" benchmark only which favor AMD FX procssors
Quote:

Originally Posted by White Fire

This is virtually useless info.

Nobody judges the game based on it performance at 1024x768

What about 1280x720 ?

 

[bencher]

What a useless review.

Let's see tests at resolutions people actually play at.

 

[MakubeX]

Quote:

Originally Posted by Cloudfire777

1024x768?

Of course, they're measuring CPU performance, not GPU. Setting a high resolution would make the GPU a bottleneck.

 

[White Fire]

Quote:

Originally Posted by ThePath

Quote:

Originally Posted by Cloudfire777

1024x768?

LOL The other thread were AMD were wining was tested @ 1280x720
http://www.overclock.net/t/1363058/pc-games-hardware-crysis-3-in-the-cpu-test-amd-fx-processors-dominate-our-benchmarks

So, 1024x768 is low resolution but 1280x720 is high ?
Quote:

Originally Posted by White Fire

This is virtually useless info.

Nobody judges the game based on it performance at 1024x768

What about 1280x720 ?

I wouldn't test below 1366x768

Even then, I'd only do 1680x1050 and 1920x1080, etc

 

[Kasp1js]

Quote:

Originally Posted by jprovido

interestingly enough amd handily beats intel cpu's on the "welcome to the jungle" benchmark.

Well I wouldn't say handily, both overclocked Intel cpus are within 5fps of the 8350, I'd say that's pretty much equal.

Either way it's interesting to see such big differences depending on the scene tested.

 

[Raqe]

Here are the 1920x1080 benches:

 

[White Fire]

^ much better

And lol at the tie between the 3770k and the 8350 in the welcome to the jungle test

 

[bencher]

^^Thank you.

 

[Kasp1js]

Quote:

Originally Posted by Raqe

Here are the 1920x1080 benches:

Spoiler: Warning: Spoiler!

That's better, it would be perfect if they had also tested with lower graphical settings thus taking the average frames closer to 60fps which is what you play at, but either way nothing unexpected - cheaper cpus being beaten by more expensive ones...

 

[ThePath]

I replaced 1024x768 with 1920 x1080 benchmarks

I did not even notice that there is higher resolution benchmark. But the results are not so different. Intel is dominating in 2 out of the 3 benchmarks. In the jungle test, it is a tie between the 3570, 3770k and the FX-8350

 

[ahnafakeef]

Even an OCed 8350 gets beaten by a 3570K, sad.

But I thought one of the reviews showed that the 8350 was actually better for Crysis 3 than a 3570K?

 

[bencher]

Looks like I will be happy going with a 8320.
Quote:

Originally Posted by ThePath

I replaced 1024x768 with 1920 x1080 benchmarks

I did not even notice that there is higher resolution benchmark. But the results are not so different. Intel is dominating in 2 out of the 3 benchmarks. In the jungle test, it is a tie between the 3570, 3770k and the FX-8350

Since when is 1 fps dominating?

 

[axipher]

Quote:

Originally Posted by ThePath

I replaced 1024x768 with 1920 x1080 benchmarks

I did not even notice that there is higher resolution benchmark. But the results are not so different. Intel is dominating in 2 out of the 3 benchmarks. In the jungle test, it is a tie between the 3570, 3770k and the FX-8350

You should leave both, CPU performance would be more accurate at a lower resolution.

 

[Kuivamaa]

PClab has done dodgy benchmarks in the past.



This is from BF3 Multiplayer at close quarters. Graph indicates that a series of i5 processors, regardless of clock, perform rougly the same. Now this wouldn't be a surprise in a single player environment where GPU is the limit. But this graph is on MP and it insinuates that stock i5s produce similar minimum fps with heavily o/c ones. Maximums are similar due to GPU limitations so the average evens out. To everyone that has played BF3 on 32-64 maps it is obvious that their results are bogus. Or they simply benched in 4 player maps.

Check post human [email protected] results. Again it seems that a system with an i5 or i7, clocked or not will produce pretty much the same minimum and average (if i understood what the graph stands for that is). Which means that the limiting factor is purely the gpu. Which contradicts all those ups and downs people are experiencing since you should get a steady 30 throughout this level. Someone with a similar ivy setup care to elaborate?Do they get similar performance on that level?

 

[SoloCamo]

All of these tests simply show they are are neck and neck and/or well within the margin of error. Nothing to see here..

Well, one things bothering me, what is your source for these benches?